IEEE Ai Assisted Devex Findings

AI-assisted tools are reshaping how developers work, yet empirical research examining their impact through developer experience (DevEx) frameworks remains limited. This study investigates the adoption of AI-assisted coding tools, including GitHub Copilot and Windsurf, within software development at a Fintech organization. We conducted a mixed-methods study that combines quantitative usage telemetry (47,336 prompts, 98,797 suggestions, and a 24.23\\% acceptance rate among 21 developers over ten months) with qualitative thematic analysis of nine practitioner reflections. The analysis maps the impact of AI tools across three dimensions of DevEx: flow state, feedback loops, and cognitive load. Our findings reveal a productivity\u2013validation tension: while practitioners report accelerated routine tasks and describe AI-assisted tools as \u201cgame changers,\u201d they continue to experience substantial validation burdens related to prompt formulation, output verification, and architectural alignment. Additionally, emergent dynamics highlight AI adoption as a socio-technical transformation requiring  DevEx monitoring rather than simple tool substitution.